Summary Project 1 The premise of this project is that Mycobacteria tuberculosis (MTB) drug resistance phenotypes as measured by standard in vitro DSTs on conventional media may not encompass the full range of responses to drug therapy that affect patient outcomes. In previous work, we have identified MTB mutations that confer drug tolerance and conditional drug tolerance among “drug resistant” clinical strains. We hypothesize that patients who do not manifest early and vigorous clinical responses to treatment may be infected with strains with mutations that confer these phenotypes. If this is shown to be true, the early detection of these mutations through the use of rapid diagnostic tools which would allow clinicians to modify drug treatment to achieve better outcomes. The goals of this study are to identify bacterial genetic determinants of 1) sub-optimal patient response to TB treatment. We will further characterize strains from people who respond and do not respond to TB treatment in terms of their ability to be grown on alternate non-conventional media, their mean inhibitory concentrations on alternate media, and their transcriptional signatures. We expect to identify specific mutations and transcriptional signatures that are associated with different growth characteristics and that these will be related to antibiotic tolerance and resistance phenotypes. We will conduct this longitudinal study at two field sites in different countries (Peru and Mongolia), enrolling active TB patients whom we will follow for interim and final treatment outcomes as measured by clinical criteria. Almost all previous studies of clinical treatment outcomes have relied on microbiological assays to determine treatment response so there are no well established norms to assess treatment response that do not include microbiological results. We propose here a panel of clinical assessments designed to measure pulmonary function, inflammatory response and respiratory symptoms, all of which we expect to improve over the first two months of effective TB treatment. Once we identify and isolate TB strains from people with sub-optimal responses to treatment, we will sequence these strains and perform a genome wide association study to identify specific variants associated with these outcomes.